Carolyn Peluso, Ph.D.
Our last several blog posts have described the causes of and the solutions to cell line contamination and misidentification. Hopefully, by now you are pretty confident that your cells are exactly what you thought. So on to the next step . . . how do you know that the cell line you have chosen for your experiments is a good model for the hypothesis you are testing?
Many investigators are asking the same question, and for good reason. One group looked at eight commonly used thyroid cancer cell lines, originally derived from thyroid tumors with diverse histological characteristics, representative of their individual states of differentiation. Microarray analysis of the cell lines revealed that all eight assume a similar dedifferentiated phenotype in vitro1. Thus, these cell lines may be useful if you are studying poorly differentiated forms of thyroid cancer. However, they may prove misleading if you are looking to answer questions about highly-differentiated thyroid tumors, and you are assuming that they have maintained the phenotypic character of the tumor from which they were derived.
Another study compared the regulation of the retinoic acid receptor between an immortalized mouse Sertoli cell line (MSC-1) and primary Sertoli cells. They found that the cell line and the primary cells behave in a similar manner, indicating that for these studies at least the cell line is a good model for Sertoli cell function. When they expanded their studies to examine the immune privilege properties of Sertoli cells, they found that this is a property the MSC-1 cells do not share2. Once again, this study demonstrates that cell lines are not always a perfect match for the disease or process under examination.
Clearly, some leg work is required when picking a cell line as a model system. First, it is always good to start with cells at early passage number. In general, cell lines are more likely to lose their parent cell character the longer they remain in culture. Second, before beginning a new set of experiments, the cell lines should be tested to ensure that, under normal conditions, the feature of the cell lines you are interested in studying matches up with the relevant primary cell. If you’ve tested that the cell line normally behaves like the primary cell, then alterations in behavior observed during your experiment are likely due to your manipulations and not a quirk of the culture conditions. Cells grown in culture, and away from their in vivo environment, inevitably lose some of their in vivo character. As long as we appreciate this truth, and are diligent about performing control and proof-of-concept experiments, then cell lines will remain a valuable tool for modeling disease, and will continue to help scientists advance their research.
Next time we will take this discussion a step further, and look at how next generation sequencing is helping investigators generate better model systems for cancer research and drug discovery. So, until next time - we wish you good data and happy culturing,
ATCC Cell Biology
